Environ. Sci. Technol. 1907, 21, 777-784
(19) Aylmore, L. A. G.; Karim, M.; Quirk, J. P. Soil Sci. 1967, 103,lO-15. (20) Dudas, M. J. Environ. Sci. Technol. 1981, 15, 840-843.
Zimmermann, U.; Munnich, K. 0.;Roether, W.; Kreutz, W.; Schubach, K.; Siegel, 0. Science (Washington, D.C.) 1966,152, 346-347.
Schmalz, B. L.; Polzer, W. L. Soil Sci. 1969, 108, 43-47. Harward, H. E,; Reisenauer, H. M. Soil Sci. 1966, 101, 326-335.
Received for review February 20, 1986. Revised manuscript received September 8, 1986. Accepted April 7, 1987.
Aqueous Ozonolysis Products of Methyl- and Dimethylnaphthalenes Michael D. Gaul,” Gregor A. Junk, and Harry J. Svec Ames Laboratory, Iowa State University, Ames, Iowa 50011
rn The ozonolyses of 1-and 2-methylnaphthalene and 1,2-, 1,3-, 1,4-,and 2,3-dimethylnaphthalene were performed in dilute aqueous solution, and the resulting nonperoxidic products were extracted, concentrated, and identified. Identification of the ozonolysis products was confirmed by comparison of retention indexes and mass and infrared spectroscopy data with those from authentic samples when possible. The compound identifications were facilitated by comparisons with products from parallel ozonolysis reactions performed in n-hexane and methanol at higher concentrations. The aqueous ozonolysis of l-methylnaphthalene yielded four major products: 2-acetylbenzaldehyde due to the cleavage of two double bonds and (E)and (2)-3-(2-acetylphenyl)-2-butenal and (Z)-3-(2-acetylpheny1)propenal due to the cleavage of a single double bond. The products resulting from the ozonolysis of 2methylnaphthalene and the isomeric dimethylnaphthalenes were analogous to those formed for 1methylnaphthalene.
Introduction Current methods of water treatment using chlorination are known to produce chlorinated hydrocarbons from dissolved humic and fulvic acids and other trace organic compounds (1-3). Virtually all of these chlorinated hydrocarbons tend to persist in the environment and bioaccumulate, and many have been shown to be toxic or oncogenic. The production of chlorinated hydrocarbons is one of many problems caused by chlorination that has increased the popularity of ozonation as an alternative for water treatment. The question of ozonation byproducts is only just beginning to be addressed. There have been a few reports involving the ozonation of natural and waste water (4-8) where elevated concentrations of hydroxy aromatic compounds, alkyl phthalates, aliphatic aldehydes and acids, alkanes, and aromatic hydrocarbons were found in waste water. One problem with these reports is that the precursors to many of the compounds produced were unknown. A simpler approach than studying the complex mixtures in ozone-treated natural or waste water is to characterize incomplete ozonation products from model compounds such as aliphatic alcohols, ketones, acids (9, IO), amino acids, sugars ( I I ) , and, perhaps most importantly, benzene derivatives (12-1 7) and polynuclear aromatic hydrocarbons (18-26). Increased interest in the benzene derivatives is due to the fact that many, such as phenols, are toxic and known to be present in industrial wastes or to be produced under certain conditions with ozone. Interest in the polynuclear
* Address correspondence to this author at his present address: ~
Dow Corning Corp., Midland, MI 48640. 0013-936X/87/0921-0777$01.50/0
~~
~~
aromatic hydrocarbons (PAHs) has been generated because of the toxicity and carcinogenicity of these compounds and the possibility that ozone might promote the formation of epoxides. Epoxides of the PAHs, in general, show a tendency to be toxic and carcinogenic (27). The reaction of the simplest PAH naphthalene with ozone in water has been reported (22-26), but only three of these reports had any specific information about the mechanism or products produced. There have been no studies involving the aqueous ozonation of the methyl- and dimethylnaphthalenes despite the fact that these compounds are known contaminants of potable water supplies at trace levels (28) and they are principal constituents of crude oils and oil products (29) that regularly contaminate our water systems. Naphthalene and its methyl derivatives are also generally more soluble in water than the higher molecular weight fused-ring aromatic compounds, and consequently, their partial ozonation products are of greater concern. The mechanism shown in Figure 1 is general for the ozonolyses of naphthalene derivatives in “participating” or nucleophilic solvents, although it is written specifically for the reaction in methanol, the solvent most frequently chosen for mechanistic studies. Ozone has electrophilic character and, therefore, normally adds to the most electron-rich ring of a substituted naphthalene, i.e., the methylated ring for the methyl- and dimethylnaphthalenes. Ozone adds to the double bond via a 1,3 dipolar cycloaddition to form the unstable 1,2,3-trioxalane ring system, which rearranges to form an aldehyde or ketone and the carbonyl oxide. The zwitterionic carbonyl oxide is very reactive, and therefore, in participating solvents, such as methanol or water, the zwitterion reacts with the solvent to form a methoxy or hydroxy intermediate. The reaction sequence is then repeated for the second double bond, which generally reacts with ozone much more easily due to the lesser degree of resonance stabilization of this bond. The breaking of the second double bond yields two intermediates, an aromatic methoxy or hydroxy hydroperoxide and a two-carbon aliphatic hydroperoxide. In aqueous solution, the aromatic hydroxy hydroperoxide either decomposes to phthalaldehyde and hydrogen peroxide or cyclizes. The cyclic peroxide is isolatable (24) but easily decomposed in the presence of acid and a reducing agent to phthalaldehyde. This general mechanism appears to be operable for the ozonolysis of the isomeric methyl- and dimethylnaphthalenes in dilcte aqueous solution as described in this paper. Efforts were made to simulate the ozonation conditions of natural or waste water treatment, and an emphasis was placed on the qualitative characterizations of the secondary, nonperoxidic reaction products since these would be the compounds most likely to persist after water treatment. The particular isomeric dimethyl-
@ 1987 American Chemical Society
Environ. Sci. Technol., Vol. 21, No. 8, 1987 777
\
L
1 Flgure 1. General mechanism for ozonolysls of naphthalene in participating solvents such as methanol and water (30).
naphthalenes were chosen for two reasons. The first reason was that the alkyl substitution can have a strong effect on the orientation of the reaction and the products produced when ozone attacks an unsaturated system (30,31). The second reason was that the methyl groups can act as a structural tag, much as an isotopic label, and this should aid in the mass spectral identification of unusual products. Background knowledge for the products of the aqueous ozonolyses was obtained by reacting these compounds with ozone in n-hexane and methanol at higher concentrations. Capillary gas chromatography (GC) and capillary gas chromatography/mass spectrometry (GC-MS) were used as the primary methods of product characterization. Experimental Section Solution Preparation. The methyl- and dimethylnaphthalenes used for the ozonolysis experiments were obtained from commercial sources. l-Methylnaphthalene, 2-methylnaphthalene, 1,2-dimethylnaphthalene, and 1,3dimethylnaphthalene were obtained from Aldrich Chemical Co., Milwaukee, WI. 1,4-Dimethylnaphthalene and 2,3-dimethylnaphthalene were supplied by Pfaltz and Bauer, Inc., Stanford, CT. Four of the six compounds were used without further purification; however, 2-methylnaphthalene and 1,4-dimethylnaphthalene were purified by passage through a short column of Florisil before use. Aqueous solutions of l-methyl- and 2-methylnaphthalene were prepared in l-L Erlenmeyer flasks by dissolving approximately 18 mg of each in HPLC-grade water, supplied by J. T. Baker Chemical Co., Phillipsburg, NJ. The solutions were stirred for 48 h at 30 "C to dissolve the methylnaphthalenes and then equilibrated at 25 "C before actual concentrations of the methylnaphthalenes were measured. One-liter aqueous solutions of the dimethylnaphthalenes were prepared in the same manner as the methylnaphthalenes, except only 1.5-2.0 mg of each dimethylnaphthalene was added to solution. The crystalline 2,3778
Environ. Sci. Technol., Vol. 21, No. 8, 1987
dimethylnaphthalene was an exception. The aqueous solution containing 2,3-dimethylnaphthalene was heated to 95 "C for 2 h with stirring and then held at 50-60 "C overnight in an effort to dissolve the sample. The excess 2,3-dimethylnaphthalene was removed by filtering with a 5 Fm porosity sintered glass funnel. Reaction Conditions. The purity of the oxygen feed gas was 96.6 % or greater, and a 4-A molecular sieve drying tube was used to remove any residual moisture. The ozonater was a Model 03V9-0 ozone generator made by Ozone Research and Equipment Corp., Phoenix, AZ. A standard 500-mL gas wash bottle with a 12 mm coarse porosity gas dispersion tube was used as an ozonation vessel. The outlet of gas wash bottle was connected to an all-glass cold trap in early experiments when purging of the starting materials or products was a concern, but in latter experiments this cold trap was omitted. The output of the ozone generator was determined by iodometric titrations as described by Boelter et al. (32). The aqueous ozonations of the methylnaphthalenes were performed with 200 mL of the l-L stock solutions. The oxygen feed gas flow was 70 mL/min, and the discharge tube pressure and current were 4 psi and 0.4 A, respectively. All reactions were performed at room temperature. The reaction times were adjusted depending on the methylnaphthalene concentration and desired completeness of reaction; generally, the reactions were stopped after reaction of 95% of the starting material. Each reaction solution was flushed for about 10 s with a high flow of nitrogen to remove any residual ozone. The aqueous ozonations of the dimethylnaphthalenes were performed in the same manner as those of the methylnaphthalenes; only 300 mL of the stock solutions was used for each ozonation. The ozonations of n-hexane and methanol were performed with the same apparatus as the aqueous reactions only at higher concentrations. Approximately 2.6-2.9 mmol of the methyl- and dimethylnaphthalenes was dissolved in 150 mL of either n-hexane or methanol, and the soltltions were treated with ozone by using 0.8-1.0 L/min oxygen flow, discharge tube pressure of 4 psi, and discharge tube current of 0.7 A. Reactions were performed at room temperature, and reaction time criteria were the same as for the aqueous reactions. The reacted solutions were purged for 1min with nitrogen to remove residual ozone and then covered and allowed to stand 48-72 h. Product Isolation. The aqueous ozonation solutions of 1-and 2-methylnaphthalene were acidified with 2.0 mL of 1M hydrochloric acid and then extracted with methylene chloride. The extracts were then dried over sodium sulfate and concentrated with Kuderna-Danish evaporative concentrators as modified by Junk et al. (33). The extracts were then transferred to 5.0-mL volumetric flasks and diluted with methylene chloride and a small amount of diazomethane reagent. The ozonation solutions of the dimethylnaphthalenes were treated in a similar manner as the solutions from 1and 2-methylnaphthaIene, except that extraction volumes were 75 mL and final solution volumes were adjusted to 1.0 mL. The ozonation solutions in n-hexane and methanol were simply concentrated before analysis. In the case of nhexane, the concentrated solution was derivatized with diazomethane. The methylene chloride, along with all other organic solvents used in this investigation, was supplied by Burdick and Jackson Laboratories, Inc., Muskegon, MI, and was redistilled before use.
Table I. Products from Ozonolysis of 1-Methylnaphthalene in Water, Methanol, and n -Hexane product phthalic anhydride 2-acetylbenzaldehyde methyl 2-formylbenzoate 3-methoxyphthalide methyl 2-acetylbenzoate methyl 2-formyl-6-methylbenzoate (or isomer) methyl 2-formyl-3-methylbenzoate (or isomer) unidentified unidentified 3-methoxy-4-methylphthalide dimethyl 3-methylphthalate (Z)-3-(2-formylphenyl)-2-butenal methyl (2)-3-(2-forrnylphenyl)-2-butenoate (E)-3-(2-formylphenyl)-2-butenal methyl (E)-3-(2-formylphenyl)-2-butenoate methyl (Z)-3-[2-[(methyloxy)carbonyl]phenyl]-2-butenoate (Z)-3-(2-acetylphenyl)propenal
retention index 1310 1334 1354 1402 1414 1420 1466 1470 1479 1497 1542 1545 1566 1583 1616 1619 1622
% M yields in
H20
CHSOH
CBHII
identification data
-a
-
16
1 6 1 3 3 2 3 46 7b 4 2 2
5
-
GC, GC-MS, GC-FTIR GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC-MS GC-MS, GC-FTIR GC-MS GC-MS GC-MS, GC-FTIR GC-MS, GC-FTIR GC-MS GC-MS, GC-FTIR GC-MS GC-MS, GC-FTIR GC-MS, GC-FTIR GC-MS
17 -
-
-
-
11
8 -
-
46 5 11 -
9
5
-
3 3
-
-
-
-
Dashes indicate a yield of less than 1% for the listed product. The yield of unidentified products was estimated by assuming that the response factor for this product was equal to that of dimethyl phthalate.
Product Identification. All GC analysis for the ozonolysis products of the methyl- and dimethylnaphthalenes was performed with a Carlo Erba Fractovap 4160 gas chromatograph obtained from Erba Instruments, Inc., Peabody, MD. The capillary separations were made with a fused silica SE-54 capillary, 30 m X 0.25 mm, made by J & W Scientific, Inc., Rancho.Cordova, CA. Hydrogen carrier gas, a splitless injection mode, and oven programming of 40-220 OC at 6 deg/min with a 1-min initial hold were used. The chromatographic output for most samples was recorded with a HP3388 integrator obtained from Hewlett-Packard, Avondale, PA. Kovats retention indexes were recorded for all sample components by comparison to a C8-C22 normal hydrocarbon standard (34). The quantification of sample components was done with external standards, making corrections for the extraction efficiencies for the water reactions. The extraction efficiencies and response factors were estimated for products for which there were no authentic samples. Packed-column gas chromatographic analyses were performed with a Perkin-Elmer 3920B gas chromatograph from Perkin-Elmer, Norwalk, CT, by using either a 3% SE-30, a 5% FFAP, or a Tenax column (1 m X 3 mm). These columns were used solely for quantitative analysis of starting materials with naphthalene as an internal standard. All GC-MS analyses were performed with a Finnigan 4000 equipped with a Model 8600 GC and the Incos 2300 data system, Finnigan Corp., Sunnyvale, CA. The scanning of the mass spectrometer and data collection was controlled by a Data General Model Nova 3 computer with the Incos 2300 software. Electron ionization at 70 eV was used for all samples, and data were collected over the mass range of 45-400 atomic mass units with scan cycle times of 1s. Chemical ionization spectra were recorded for some samples with isobutane or methane as the reagent gas. Data were collected over a 75-300 mass range, again with a 1-s scan cycle time. The gas chromatographic column used was a fused silica DB-5, with the same temperature programming described earlier and helium carrier gas. The GC and GC-MS retention data were correlated with an easily characterized major component as a reference compound. Capillary gas chromatography-Fourier transform infrared spectrometry (GC-FTIR) analyses were also performed on selected product mixtures with a Bruker-IBM
Model 98 Fourier transform infrared spectrometer obtained from IBM Instruments, Inc., Danbury, CT. The instrument was linked to a Hewlett-Packard 5800 gas chromatograph equipped with a thick film (1.0 pm) fused silica DB-5 capillary column, 30 m X 0.25 mm. The gas chromatographic conditions were the same as those described previously. The scan rate of the instrument was approximately 3.4 scans per second with 8-cm-l resolution. Data were collected over the range of 4000-800 cm-l. Synthesis of Authentic Standards. An authentic standard of 3-methoxy-3-methylphthalide was synthesized from 2-acetylbenzoic acid by refluxing in HC1-acidified methanol. Some of the properties of 3-methoxy-3methylphthalide have been reported in the literature (35). A mixture of cis- and trans-1,3-dimethoxyphthalan was made from phthalaldehyde and 3-methoxyphthalide was made from 2-formylbenzoic acid by the same procedure as that described for 3-methoxy-3-methylphthalide. Results and Discussion Product Characterization. Tables I-IV list the name, retention index, molar yields, and identification data of the ozonolysis products for each of the methyl- and dimethylnaphthalenes in water, methanol, and n-hexane. All the products characterized from the six aqueous ozonolysis reactions were aldehydes and ketones. No acids were characterized. Determination of the extraction efficiencies of aromatic acid products, such as 2-acetylbenzoic acid and 2-formylbenzoic acid, at 1 ppm concentrations indicated that these products would have been isolated if present. Previous studies (24-26) involving the aqueous ozonolyses of naphthalene had shown increased yields of phthalaldehyde in comparison with reactions in organic solvents, but not to the total exclusion of the aromatic acid products. Acidic products would be expected from the oxidation of corresponding aldehydes by peroxides present, particularly hydrogen peroxide formed by elimination from the hydroxy hydroperoxide precursor as shown in Figure 1. Ozone should play a minor role in the formation of acids because it is relatively unreactive toward aldehydes (30). Apparently little or no oxidation of the aldehydes occurred, probably because the solutions were 2.5-40 times more dilute than in previous work (26) and product residence times in the water were short. The products resulting from the cleavage of two double bonds in the naphthalene ring system were expected on Environ. Scl. Technol., Voi. 21, No. 8, 1987
779
Table 11. Products from Ozonolysis of 2-Methylnaphthalene in Water, Methanol, and n -Hexane
retention index
product phthalaldehyde phthalic anhydride (E)-and (2)-1,3-dimethoxyphthalan
1238 1310 1318 1321 1354 1360 1402 1455 1474 1479 1517 1520 1530 1554 1569 1574 1601 1612 1621 1658 1664
methyl 2-formylbenzoate 2-formyl-4-methylbenzaldehyde 3-methoxyphthalide dimethyl phthalate methyl 2-formyl-5-methylbenzoate (or isomer) methyl 2-formyl-4-methylbenzoate (or isomer) (2)-2-(3-oxo-l-butenyl)benzaldehyde 3-methoxy-6-methylphthalide (or isomer) 3-methoxy-5-methylphthalide (or isomer) (2)-3-(2-formylpheny1)-2-methylpropenal (E)-3-(2-formylphenyl)-2-methylpropenal dimethyl 4-methylphthalate (2)-3-(2-formyl-5-methylphenyl)propenal (or isomer) (2)-3-(2-forrnyl-4-methylphenyl)propenal(or isomer) (E)-2-(3-oxo-l-butenyl)benzaldehyde (E)-3-(2-formyl-5-methylphenyl)propenal(or isomer) (E)-3-(2-formyl-4-methylphenyl)propenal (or isomer)
% M yields in
HzO
CHBOH
C6Hll
identification data
20
10
12 6 -
GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR
14
GC, GC-MS, GC-FTIR GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC-MS, GC-FTIR GC-MS, GC-FTIR GC-MS GC-MS, GC-FTIR GC-MS, GC-FTIR GC-MS GC-MS GC-MS, GC-FTIR GC-MS GC-MS GC-MS GC-MS GC-MS
-
-a
-
9
-
16 -
4
9 2 2
14 3 5 5 5
5
1 -
2 2 2 2 2
-
-
3
4 4
-
-
-
'Dashes mean a yield of less than 1%for listed product. Table 111. Products from Ozonolysis of 1,2-Dimethylnaphthalene in Water, Methanol, and II -Hexane
retention index
product
(E)-or (Z)-l-methyl-1,3-dimethoxyphthalan phthalic anhydride 2-acetylbenzaldehyde 3-methoxy-3-methylphthalide 3-methoxyphthalide methyl 2-acetylbenzoate unidentified methyl (2)-3-(2-formylphenyl)-2-methyl-2-butenoate methyl 2-formyl-3,4-dimethylbenzoate (or isomer) (2)-3-(2-formylphenyl)-2-methyl-2-butenal 3-methoxy-6,7-dimethylphthalide (or isomer) methyl (E)-3-(2-formylphenyl)-2-methyl-2-butenoate (2)-1-(2-acetylphenyl)-l-buten-3-one
1296 1310 1334 1381 1402 1414 1479 1611 1617 1618 1647 1685 1686
% M yields in
H2O
CHBOH
C6Hi.l
identification data
-a
2 2 11 12 7 6
-
GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC-MS GC-MS GC-MS GC-MS GC-MS GC-MS GC-MS
-
65
-
-
-
-
32 -
-
9* 2 -
3 6 -
2
8
-
27
-
-
3 4
2
Dashes mean a yield of less than 1% for the listed product. The yield of unidentified products was estimated by assuming that the response factor for this product was equal to that of dimethyl phthalate. Table IV. Products from Ozonolysis of 1,3-Dimethylnaphthalene in Water, Methanol, and n -Hexane
product (E)-and (Z)-l-methyl-l,3-dimethoxyphthalan 2-acetylbenzaldehyde 3-methoxy-3-methylphthalide 3-methoxyphthalide methyl 2-acetylbenzoate (21-24 1-methyl-3-oxo-1-buteny1)benzaldehyde methyl 2-formyl-4,6-dimethylbenzoate (or isomer) 3-methoxy-4,6-dimethylphthalide (or isomer) dimethyl 3,5-dimethylphthalate methyl (Z)-2-(l-methyl-3-oxo-l-butenyl)benzoate (Z)-3-(2-acetylphenyl)-2-methylpropenal
retention index 1279 1296 1334 1381 1402 1414 1547 1578 1606
1611 1616 1637
% M yields in
H2O -a
27 -
21
CH3OH 8 11 12 4
8 7 3
-
-
11
1 1 2
-
CBHli -
14 -
26 4 3 -
-
identification data GC-MS, GC-FTIR GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC-MS GC-MS GC-MS GC-MS GC-MS GC-MS
"Dashes mean a yield of less than 1%for the listed product.
the basis of previous literature results. The aromatic products 2-acetylbenzaldehyde and phthalaldehyde from 1-and 2-methylnaphthalene were strong evidence that the normal ozonolysis mechanism described in Figure 1 had occurred. These two aldehydes would have been expected from the decomposition of their respective hydroxy hydroperoxides. As predicted from the mechanism, the 780
Environ. Sci. Technol., Vol. 21, No. 8 , 1987
normal ozonolysis products for the dimethylnaphthalenes were observed to be 2-acetylbenzaldehyde, phthaldehyde, 2-formyl-4-methylbenzaldehyde, and l,Bdiacetylbenzene, as shown in Tables 111-VI. The predominate products for the aqueous ozonolyses were from the cleavage of only one double bond in the naphthalene ring system. Previous literature (30,36,37)
Table V. Products from Ozonolysis of 1,4-Dimethylnaphthalene in Water, Methanol, and n -Hexane
product
( E )- and (Z)-1,3-dimethyl-1,3-dimethoxyphthalan 3-methoxy-3-methylphthalide 1,2-diacetylbenzene methyl 2-acetylbenzoate methyl 2-formyl-3,6-dimethylbenzoate methyl (Z)-3-(2-acetylphenyl)-2-butenoate (Z)-3-(2-acetylphenyl)-2-butenal (E)-3-(2-acetylphenyl)-2-butenal
retention index
% M yields in
HzO
CHSOH
CBH,,
identification data
3 5
1328 1339 1381 1399 1414 1538 1585 1598 1639
GC-MS GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC-MS, GC-FTIR GC-MS, GC-FTIR GC-MS, GC-FTIR GC-MS
11 21
4
-
3
-
ODashes mean a yield of less than 1% for the listed product. Table VI. Products from Ozonolysis of 2,3-Dimethylnaphthalene in Water, Methanol, and n -Hexane
product phthalaldehyde phthalic anhydride (E)-and (Z)-1,3-dimethoxyphthalan methyl 2-formylbenzoate 3-methoxyphthalide dimethyl phthalate and methyl 2-formylbenzoate dimethyl acetal (Z)-2-(2-methyl-3-oxo-l-butenyl)benzaldehyde 2-formyl-4,5-dimethylbenzaldehyde methyl 2-formyl-4,5-dimethylbenzoate (E)-2-(2-methyl-3-oxo-l-butenyl)benzaldehyde 3-methoxy-5,6-dimethylphthalide dimethyl 4,5-dimethylphthalate
retention index 1238 1310 1318 1321 1354 1402 1455 1549 1557 1627 1638 1680 1720
% M yields in
H20
CH30H
4
7
CBH14
GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC, GC-MS, GC-FTIR GC-MS GC-MS GC-MS GC-MS GC-MS, GC-FTIR GC-MS
-n
-
-
14
-
--
17 18 206
27
-
5
8 1 -
-
5c -
-
12 6
-
-
5 3
identification data
27 15
Dashes mean a yield of less than 1% for the listed product. *Total yield of both compounds. Total yield of dimethyl phthalate only.
indicated that these products were usually not isolated after the ozonolyses of naphthalene derivatives in organic solvents because the second double bond, which is olefinic, was expected to be more reactive than the first double bond, which is aromatic. The only exception to this generalization was the reaction of 1-and 2-naphthol (38,39), where the hydroxy group increased the rate of ozonolysis of the fiist double bond relative to the second to the extent that reaction rates were competitive. The only other report of monoozonolysis products being characterized was by Kruithof and Heertjes (25) for the aqueous ozonolysis of naphthalene, although only very small amounts of the monoozonolysis product were characterized early in the reaction sequence before most of the naphthalene had been consumed. Some of the specific compounds characterized in this investigation have not been reported in the literature. The structure elucidation of these compounds is based predominantly on interpretation of the E1 and CI mass spectra. Supplementary information was obtained from carbonyl infrared absorption frequencies for products formed in n-hexane and methanol. The ozonolysis of 2-methylnaphthalene in water produced all four possible monoozonolysis products resulting from a double-bond cleavage in either the methylated or unmethylated ring. The major product resulted from bond cleavage in the methylated ring, and Figure 2 shows the two structural possibilities with their expected base peak m / z values. The base peak of this product at m/z 131 in the E1 spectrum and the pseudo molecular ion in the CI spectrum at m/z 175 clearly indicate that an acetyl radical is lost to form the base peak, and therefore, the upper structure in Figure 2 is the only possible structure. The E1 and CI mass spectra for the remaining three products were very similar, and therefore, structural assignments
--+M/Z
131
+ M/Z
145
kCOCH3
- ',
C HO
Figure 2. Two posslble monoozonolysis products resulting from oxidation of the methylated ring of 2-methylnaphthalenealong with the base peak m l r values from each compound's E1 mass spectrum.
were made on the basis of minor differences in the GC retention indexes. The ozonolysis of l-methylnaphthalene in water could also produce two possible monoozonolysis products resulting from bond cleavage in the methylated ring. Unfortunately, neither the base peak fragmentation nor the CI mass spectra could distinguish between the two possible products; however, evidence gathered from the reaction of l-methylnaphthalene in methanol indicated that the major product in water was (2)-3-(2-formylpheny1)-2butenal. The monoozonolysis product from the aqueous ozonolysis of 1,2-dimethylnaphthalene eluted at a retention index of 1618. The base peak at m/z 159 in the E1 mass spectrum indicated that this product resulted from the cleavage of the bond between carbons 3 and 4. This cleavage is consistent with the predominate product found for l-methylnaphthalene; however, in the case of 1,2-dimethylnaphthalene the alternate product resulting from bond cleavage between carbons 1 and 2 was not found. The product mixture resulting from the aqueous ozonolysis of 1,3-dimethylnaphthalenecontained both possible monoozonolysis products resulting from the oxidation of Environ. Sci. Technol., Vol. 21, No. 8, 1987
781
the methylated ring. Once again, the base peak distinguished between the two products. Finally, the aqueous ozonolysis of the symmetrical 1,4and 2,3-dimethylnaphthalene yielded only one monoozonolysis product for each as indicated in Tables V and VI, respectively. The geometrical isomers of many of the monoozonolysis products were sometimes found in the product mixtures, as indicated in Table I. The expected Configuration around the double bond of the monoozonolysis products is Z; i.e., the phenyl substituent and carbonyl-containing substituent are cis to one another. Unfortunately, the acidification of the water solutions prior to extraction would be expected to promote partial isomerizations of these double bonds by protonation of the oxygen in the carbonyl substituent. There was no direct spectral evidence to differentiate the 2 and E isomers of a particular product. However, the assumptions were made that the E isomer was lower in concentration, had a similar mass spectrum, and a later GC retention time than the 2 isomer. When the E isomers appeared, there was usually little difficulty in making structural assignments. It is interesting that the monoozonolysis products were predominant for the aqueous reactions. Apparently, the second double bond to be broken during ozonolysis of the naphthalene ring system is very unreactive toward ozone when the ozonolysis is performed in water. This unreactivity toward ozone may be due to steric interactions promoted by the strong hydration of the polar functionality formed during the course of the reaction. When the first double bond is broken, the remaining double bond and its polar substituents can assume any one of several configurations with respect to the benzene ring. The configuration depends on both intramolecular steric interactions and intermolecular interactions, such as hydration of the polar functionality. The configuration of this double bond and the degree of hydration of its polar substituents will have a large effect on the bond's reactivity toward ozone, since the ozone molecule normally requires an orthogonal access to the double bond. The second double bond would be expected to have greater configurational freedom in organic solvents, including methanol, where solvation occurs to a smaller degree. The product characterizations in n-hexane and methanol were consistent with this idea since the monoozonolysis products were isolated in these solvents as only minor components of the product mixtures. The product yields for the six ozonolysis reactions are given in Tables I-VI. For many of the products, particularly the monoozonolysis products, these yields are not exact because there were no authentic samples from which to determine accurately the response factors and extraction efficiencies. The response factors for monoozonolysis products were estimated to be the same as that found for cinnamaldehyde, or 1.5 relative to dimethyl phthalate. The relative response factors for products resulting from the cleavage of two double bonds, such as methyl 2-formylbenzoate (0.94), phthalaldehyde (1.05), and 2-acetylbenzoate (1.13), were predictably lower than those found for the monoozonolysis products due to the lower carbon-hydrogen content. Another factor that could have affected product yields was the purging of either starting material or product from solution during the reactions. Experiments using 100 mL/min flow rates of just oxygen without ozone generation indicated that less than 5% of the starting materials were purged from water solution during the time of ozonolysis reactions. The most volatile product phthaldehyde had 782
Environ. Sci. Technol., Vol. 21,
No. 8, 1987
7;
Po m
I I/
T IME Figure 3. Chromatograms from the gas chromatographic analysis of product mixtures resulting from the ozonolyses of 1-methylnaphthalene in water with 4 (top), 12 (middle), and 48 (bottom) molar equiv of ozone.
no tendency to be purged from solution. One final note conerning the effect of large excesses of ozone on reaction yields is as follows. Ozone is a powerful oxidant, and it will continue to react with the primary products, the ortho-substituted benzenes, after the methylor dimethylnaphthalene has been consumed. The primary products are presumably degraded to short-chain, difunctional aldehydes, ketones, and acids and eventually to carbon dioxide and water. Indirect evidence that this occurred under the experimental conditions used in this investigation is demonstrated with the GC product profiles shown in Figure 3. The top profile in Figure 3 results from aqueous ozonolysis of 1-methylnaphthalene with only 4 molar equiv of ozone, an insufficient amount for complete reaction. The middle and bottom GC profiles represent product mixtures resulting from the addition of 12 and 48 molar equiv of ozone, respectively,to 1-methylnaphthalene. With use of product 1334 to compare the profiles, it becomes obvious that the concentrations of all products decrease dramatically with increasing amounts of ozone beyond that necessary for complete reaction of the starting material. Although the ortho-substituted benzene products would normally be thought of as quite resistant to further oxidation, the addition of ozone had to be closely controlled to prevent the degradation of these products and the resulting reduction of reaction yields based on these products. In general, the isolated product yields were reasonably good. Total product yields were greater than 50%, not including some minor unidentified products isolated for some of the product mixtures, notably that obtained from the reaction of 2,3-dimethylnaphthalene.The expected mechanism of ozonolysis suggests that the unaccounted molar yield of 20-50% for the six reactions is probably due to hydroxy peroxides, which are stable to acid but not volatile or stable enough to be gas chromatographed. Support for this suggestion comes from the results of qualitative tests for peroxides, which indicated their
3500
2500
c Jj
,
2000
'
16DO
I200
E
C M' Figure 4. GC-FTIR spectra of 3-methoxyphthalide(top) and methyl 2-formylbenzoate (bottom).
presence, and previous literature (24-26), which indicated that the aqueous ozonolysis of naphthalene yields approximately 45% of the stable cyclic hydroxy peroxide and 55 76 phthalaldehyde. Ozonolysis in n-Heaaae and Methanol. The following paragraphs provide some general discussion concerning the products of the ozonolysis reaction performed in n-hexane and methanol. The product mixtures resulting from the reactions in n-hexane were quite different than those mixtures obtained for the ozonolyses performed in water. The biggest difference was the presence of the carboxylic acids in the product mixtures from n-hexane. The acids were in all cases the more highly oxidized form of the aldehydic products observed in the aqueous ozonolyses. For example, the reaction of 1-methylnaphthalene yielded 2-acetylbenzaldehyde and methyl 2-acetylbenzoate (after derivatization) as major products. Small amounts of the expected ring-methylated products 2-formyl-2-methylbenzaldehyde and methyl 2-formyl-3-methylbenzoatewere also characterized. The expected monoozonolysis product (2)-3-(2-formylphenyl)-2-butenal was characterized, but its more highly oxidized analogue was not. The presence of the acidic products was probably due to the slightly different mechanism that exists in n-hexane for the ozonolysis. The peroxide intermediates that are produced during the ozonolysis in n-hexane are believed to be predominantly polymeric (30) and insoluble in the nonpolar solvent. The evidence for this was the observation of a white flocculant solid that precipitated during the course of the reaction. These polymeric peroxides then decompose in a mafiner different from the intermediates obtained in participating solvents, such as water and methanol, and this decomposition can yield carboxylic acid substituted products directly. Experimental evidence for the oxidation of the n-hexane solvent also suggests that
high concentrations of aliphatic peroxides might be responsible for the oxidation of aldehydic functionality associated with the aromatic products. Due to this peroxidic environment, the product mixtures in n-hexane were found to be unstable with time. Another basic difference between the product mixtures in organic solveilts and water was the fact that the monoozonolysis products were only minor products for the reactions in n-hexane and methanol and were completely absent in the product mixtures resulting from the reaction of 2-methylnaphthalene and 2,3-dimethylnaphthalene in the two organic solvents. In all cases, the molar yield of the monoozonolysis products was less than lo%, and the relative amounts of the monoozonolysis products, compared to those of products resulting from the cleavage of both double bonds, were dramatically different from that found for the aqueous reactions. For the reactions in organic solvents, the products resulting from the cleavage of both double bonds greatly predominated. The ozonolysis product mixtures in methanol were much more complex than the product mixtures obtained in water or in n-hexane. The products obtained in methanol were those expected from the previous literature (30);however, the mixtures were complicated by the solvent's ability to promote the formation of stable cyclization products, such as acetals, ketals, and phthalides. One of the best examples of this was the product mixture resulting from the reaction of %methylnaphthalene in methanol. Not only were the expected products phthalaldehyde and methyl 2-formylbenzoate and the isomeric methyl 2-formyl-4(or 5)methylbenzoates characterized, but their cyclized analogues (E)-and (2)-1,3-dimethoxyphthaIan and 3-methoxyphthalide and the isomeric 3-methoxy-5(or 6)methylphthalides were also found. GC-FTIR was instrumental in confirming the structure of many of the products formed in methanol. Isomeric products such as methyl 2-formylbenzoate and 3-methoxyphthalide were difficult to discriminate on the basis of their mass spectra, mainly because the molecular ions and many of their fragment ions were the same, varying only in intensity. Positive strhcture identification, however, can easily be made by comparing their infrared spectra as shown in Figure 4. The substituted benzoate had two Carbonyl adsorptions at 1743 and 1716 cm-l and the phthalide had a single adsorption at 1813 cm-l. Many compounds characterized in this study were ideal candidates for GC-FTIR analysis. Few organic functional groups are IS strongly absorbihg in the infrared region as the carbonyl; thus, sensitivity was excellent for carbonyl or carbon-oxygen single-bond adsorptions. In most cases, the number of and the wavelength for the carbonyl adsorptions were all that were required to confirm a structure already partially elucidated by GC-MS. A library of GCFTIR spectra (along with the mass spectra) for each of the products and discussions of many interesting steric and electronic effects on the carbonyl adsorption frequencies is published elsewhere (40). Acknowledgments We thank W. S. Trahanovsky for his interest in the project and Mike Avery, John Richard, and Steve Veysey for instrumental assistance and general advice. Registry No. 1-Methylnaphthalene, 90-12-0; 2-acetylbenzaldehyde, 24257-93-0; methyl 2-formyl-6-methylbenzaate, 63112-99-2; methyl 2-formyl-3-methylbenzoate, 108293-41-0; (2)-3-(2-formylphenyl)-2-butenal, 108293-42-1;methyl (2)-3-(2formylphenyl)-2-butenoate, 108293-43-2; (E)-3-(2-formylphenyl)-2-butenal, 108293-44-3;(Z)-3-(2-acetylphenyl)propenal, Environ. Scl. Technol., Vol. 21, No. 8, 1987
783
108293-45-4; 2-methylnaphthalene, 91-57-6; phthalaldehyde, 643-79-8; phthalic anhydride, 85-44-9; (E)-l,3-dimethoxyphthalan, 22882-31-1; (2)-1,3-dimethoxyphthalan,22882-30-0; methyl 2formylbenzoate, 4122-56-9; methyl 2-formyl-5-methylbenzoate, 108293-46-5; methyl 2-formyl-4-methylbenzoate, 63112-98-1; (2)-2-(3-oxo-l-butenyl)benzaldehyde, 108293-47-6;3-methoxy6-methylphthalide, 108293-48-7;3-methoxy-5-methylphthalide, 63113-02-0; 1,2-dimethylnaphthalene,573-98-8; 3-methoxy-3methylphthalide, 1077-59-4; 3-methoxyphthalide, 4122-57-0; methyl 2-acetylbenzoate, 1077-79-8; (2)-3-(2-formylpheny1)-2methyl-2-butenal, 108293-49-8;methyl (E)-3-(2-forrnylphenyl)2-methyl-2-butenoate, 108293-50-1; 1,3-dimethylnaphthalene, 575-41-7; (E)-l-methyl-1,3-dimethoxyphthalan, 60026-77-9; (Z)-l-methyl-1,3-dimethoxyphthalan, 60026-76-8; (2)-2-(1methyl-3-oxo-l-butenyl)benzaldehyde, 108293-51-2; (2)-3-(2acetylphenyl)-2-methylpropenal,108293-52-3; 1,4-dimethylnaphthalene, 571-58-4; (2)-1,3-dimethyl-l,3-dimethoxyphthalan, 108293-53-4; 1,2-diacetylbenzene, 704-00-7; (2)-3-(2-acetylphenyl)-2-butenal, 108293-54-5;2,3-dimethylnaphthalene,58140-8; dimethyl phthalate, 131-11-3; methyl 2-formylbenzoate dimethyl acetal, 87656-31-3; (Z)-2-(2-methyl-3-oxo-l-butenyl)benzaldehyde, 108293-55-6;2-formyl-4,5-dimethylbenzaldehyde, 25445-42-5; methyl 2-formyl-4,5-dimethylbenzoate, 108293-56-7; 3-methoxy-5,6-dimethylphthalide, 108293-57-8.
Literature Cited Proceedings of the Conference on the Environmental Impact of Water Chlorination; Jolley, R. L., Ed.; NTIS, USDC: Springfield, VA, 1976; CONG-751096, all papers therein. Proceedings of the Second Conference on the Enuironmental Impact of Water Chlorination; Jolley, R. L.; Gorchev, H.; Hamilton, D. H., Jr., Eds.; Ann Arbor Science: Ann Arbor, MI, 1978; all papers therein. Proceedings of the Third Conference on Water Chlorination: Environmental Impact and Health Effects: Jolley, R. L.; Brungs, W. A.; Cumming, R. B.; Jacobs, V. A., Eds.; Ann Arbor Science: Ann Arbor, MI, 1980; all papers therein. Sievers, R. E.; Barkley, R. M.; Eiceman, G. A.; Shapiro, R. H.; Walton, H. F.; Dolonko, K. J.; Field, L. R. J . Chromatogr. 1977, 142, 745-754. Chappell, W. R.; Sievers, R. E.; Shapiro, R. H. Project Summary: The Effect of Ozonation of Organics in Wastewater; Health Effects Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency: Cincinnati, OH, Feb. 1981; EPA-600/S1-
81-005. Lawrence, J.; Tosine, H.; Onuska, F. I.; Comba, M. E. Ozone: Sci. Eng. 1980, 2, 55-64. J. Mallevialle, In Proceedings of the 2nd International Symposium 00 Ozone Technology, Montreal, 1975; Rice, R. G.; Pichet, P.; Vincent, M.; Eds.; Ozone Press Internationale: Jamesville, NY, 19’75; pp 133-141. Shapiro, R. H.; Kolonko, K. J.; Binder, R. T.; Barkley, R. M.; Eiceman, G. A.; Haack, L. P.; Sievers, R. E. In Proceedings of the Joint Conference on Sensing Environmental Pollutants, 4th, New Orleans, LA, 1977; American Chemical Society: Washington, DC, 1978; pp 507-510. Niki, E.; Yamamoto, Y.; Saito, T.; Nagano, K.; Yokoi, S.; Kamiya, Y. Bull, Chem. SOC. J p n . 1983,56, 223-228. Sturesund, H. J.; Bernatek, E. Acta Chem. Scand. 1970, 24, 3237-3242. Yamada, H.; Somiya, I. Ozone: Sci. Eng. 1980,2,251-260. Legube, B.; Lqnglais, B.; Dore, M. Prog. Water Technol. 1980,12, 553-570. Gilbert, E. Water Sci. Technol. 1982, 14, 849-861.
784
Environ. Sci. Technol., Vol. 21, No. 8, 1987
(14) Pringle, W. NTIS Report PB-81-115651; NTIS: Middletown, CT, Sept. 1980. (15) Dore, M.; Langlais, B.; Legube, B. Water Res. 1978, 12, 413-425. (16) Legube, B.; Langlais, B.; S o b , B.; Dore, M. Ozone Sci. Eng. 1981,3, 33-48. (17) Leitis, E. NTIS Report PB80-174394; NTIS: Springfield, VA, Feb. 1980. (18) Helleur, R.; Malaiyandi, M.; Benoit, F. M.; Benedek, A. Ozone: Sci. Eng. 1979, I, 249-261. (19) Sturrock, M. G.; Cline, E. L.; Robinson, K. R. J. Org. Chem. 1963,28, 2340-2343. (20) Chen, P. N.; Junk, G. A.; Svec, H. J. Enuiron. Sci. Technol. 1979, 13, 451-454. (21) Meineke, I.; Klamberg, H. Fresenius’ 2. Anal. Chem. 1978, 293, 201-204. (22) Meineke, I.; Klamberg, H. Fresenius’ 2. Anal. Chem. 1978, 293,205-207. (23) Neely, W. C. NTIS Report PB82-254970; NTIS: Springfield, VA, Jan. 1981. (24) Sturrock, M. G.; Cravy, B. J.; Wing, V. A. Can. J. Chem. 1971,49, 3047-3051. (25) Kruithof, J. C.; Heertjes, P. M. In Chemistry in Water Reuse; Cooper, W. J.; Ed.; Ann Arbor Science: Ann Arbor, MI, 1981; Chapter 19. (26) Legube, B.; Guyon, S.; Sugimitsu, H.; Dore, M. Water Res. 1986,20, 197-208. (27) Weil, C. S.; Condra, N.; Haun, C.; Striegel, J. A. Am. Znd. Hyg. ASSOC.J. 1963, 24, 305-325. (28) Junk, G. A.; Stanley, S. E. Ames Laboratory Report IS-3671; Ames Laboratory: Ames, IA, July 1975. (29) Neff, J. M. Polycyclic Aromatic Hydrocarbons in the Aquatic Environment. Sources, Fates, and Biological Effects; Applied Science: London, 1979. (30) Bailey, P. S. Ozonation in Organic Chemistry; Academic: New York, 1982; Vol. 11. (31) Bailey, P. S. Ozonation in Organic Chemistry; Academic: New York, 1978; Vol. I. (32) Boelter, E. D.; Putnam, G. L.; Lash, E. I. Anal. Chem. 1950, 22, 1533-1535. (33) Junk, G. A.; Richard, J. J.; Gieser, M. D.; Witiak, D.; Witiak, J. L.; Arguello, M. D.; Vick, R.; Svec, H. J.; Fritz, J. S.; Calder, G. V. J. Chromatogr. 1974, 99, 745-762. (34) Stoodley, K. D. C.; Lewis, T.; Stainton, C. L. S. Applied Statistical Techniques; Ellis Horwood: West Sussex, England, 1980; Chapter 2. (35) Erley, D. S.; Potts, W. J.; Jones, P. R.; Desio, P. J . Chem. Ind. (Moscow) 1964, 1915-1916. (36) Wibaut, J. P.; Van Dijk, J. Recl. Trau. Chim. Pays-Bas 1946, 65, 413-426. (37) Callighan, R. H.; Wilt, M. H. J. Org. Chem. 1961, 26, 5212-5214. (38) Bernatek, E.; Frengen, C. Acta Chem. Scand. 1962, 16, 2421-2428. (39) Bailey, P. S.; Johnson, C. D. J . Org. Chem. 1964, 29, 703-707. (40) Gaul, M. D. Ph.D. Dissertation, Iowa State University, Ames, IA, 1984. Received for review October 6,1986. Accepted April 6,1987. This work was supported by National Science Foundation Grant CHE 7906108. The facilities of the Ames Laboratory, US.Department of Energy, were used with support by the Pollutant Characterization and Safety Research Division, Office of Health and Environmental Research, Office of Energy Research.
